The resulting statistics of recurrent melanoma indicate a dearth of effective treatments for this disease. In order to identify targets for improved therapy, and to better elucidate the mechanisms underlying the pathogenesis of this disease, we are performing gene expression profiling of human melanoma biopsies. Using a technique known as the serial analysis of gene expression (SAGE), which allows for the quantification of individual transcripts within a cell, as well as identifies novel transcripts, we can study the gene expression profiles of melanomas at different stages in this disease using minute amounts of tissue. Currently, available on the public databases (http://cgap.nci.nih.gov) are three libraries, which we generated from three melanoma tissues, representing two vertical growth phase tumors, and one distant metastasis. In progress are libraries from compound nevi, more metastases and melanocytes. By using analysis most often reserved for microarray studies, we can examine these data using tools rarely used to analyze SAGE data, and data from a small number of SAGE libraries can be expanded using microarray analysis. These types of multilevel analyses can reveal which genes underlie progression and which genes might make viable targets for future therapy. Thus far, data generated by microarray analysis as well as SAGE have highlighted the importance of G-protein mediated signaling, resulting in the activation of PKC and rises in intracellular calcium in melanoma progression. These effects can be mediated by the gene WNT5A, the over expression of which can lead to increases in melanoma cell motility and invasion. RNAi inhibition of this pathway, followed by microarray analysis, reveals that Wnt5a may make this contribution to invasion by silencing the expression of metastasis suppressers such as Kiss-1 and NME-1. In addition, Wnt5a can suppress the expression of melanoma antigens, implying it may play a role in escaping immune surveillance. Recent data suggests that this occurs via the activation of STAT3. Another gene implicated by the gene expression data is the gene Claudin 1. We have found that Claudin 1 is highly but aberrantly expressed in melanoma, and that it is also regulated via PKC. Furthermore, increased expression of claudin 1 results in increases in invasion in melanoma cells, with concomittant increases in the matrix metalloproteinase enzymes MMP9 and MMP2. Staining of a tissue microarray indicates that this protein is upregulated in melanoma as compared to benign nevi. This gives both an upstream regulator of PKC (Wnt5a) as well as a downstream target of this enzyme (claudin 1), both of which contribute to increased invasiveness in melanoma. It is hoped that identifying important pathways such as these with high throughput gene expression profiling techniques will lead to identifying better molecular targets for treating this disease.